When the world sings the blues...

...does it mean everybody is sad? Humans can pick out 4 million shades, but different nationalities ascribe different properties to various hues. Anne M. Sebba meets the scientists who are trying to develop a world scale of colour measurement.

Anyone who has ever tried ordering clothes or household items over the internet only to find when the object arrives that the colour is not the same as it appeared on screen, knows how frustrating - and tantalising - e-commerce can be.

But when you consider that the human eye can pick out some 4 million subtle variations in colour, your frustration will give you some appreciation of the nature of the task facing scientists trying to develop a new, internationally recognised quantitative scale of colour measurement.

"It's the internet that is driving us," says James Nobbs, senior lecturer in the department of colour chemistry at Leeds University and one of just 300 colour chemists in the world.

Next month he will address colleagues from Thailand, Spain, Hong Kong and Japan at the Kyoto Institute of Technology. The occasion is the world's first conference on colour emotion, which aims to determine whether the aesthetic properties of colours are measurable.

"The ease of generating colour images on television and computer screens as well as inkjet printers has meant that the demand for people who are not colour specialists to match and describe colour has become immense. The major players in the computer industry are saying, 'If you do not come up with a new method, we will do the research ourselves.' There is a real danger that they will take over our subject.

"In the past, the demand to measure colour came only from the textile, paint and plastics industries, who were mostly happy to rely on samples and pattern books," Nobbs says.

"But selling globally on the internet means that manufacturers have to be aware of whether something is a good match to something else when other factors, such as background and texture, are taken into consideration; for example, if you are adding a button to a dress or upholstery in a car, does this look the same to everyone's eyes?" Nobbs, chairman of the technical committee appointed by the Vienna-based CIE (Commission Internationale de l'Eclairage) to devise a new independent colour scale, is under pressure to have a system in place within two years, a timetable he calls "ambitious", especially as his research, using a mixture of physics and psychology, has thrown up some unexpected results.

Culture, environment and landscape are just some of the factors that make different nationalities describe colours in different ways. After all, why do we avoid blue food, why are we told sending letters in green ink is bad form, why is red considered a good colour to wear at an interview and black favoured in some countries for mourning? These are just some of the age-old emotions and stimuli that tell the brain how to interpret colour.

The first attempt to develop a systematic way of understanding the relationship of one colour to another was made in 1905 by A. H. Munsell, an artist whose worldwide atlas was based on a simple system of paint chips with a central axis going from black to white.

Twenty-five years later, physicists devised a system whereby colour was measured with instruments according to the amount of light reflected in each wavelength. A given colour will belong to a point on the scale between light and dark, and the eye will locate it in terms of the stimulus to its red, green and blue-sensitive light receptors. This remained the standard method adopted by the CIE until 1976, when it was revised to allow for more psychological interpretations.

"It's been difficult to move away from the numerical system because psychological aspects of describing colour are complex. Even simple terms such as strong or weak, light and dark can be controversial," Nobbs says.

To prove his point he hands me a simple blue and yellow box from his desk and asks me to describe both colours in terms of their lightness and strength.

Having agreed that the blue is coloured and light, he goes on to explain that a Japanese person would use quite different terms to describe the colours - "which could have disastrous results if a Japanese carmaker asked for a lighter blue because a British paintmaker could make it by going in completely the wrong direction. I do not believe it is just a question of semantics - although the Japanese language does have more words for describing colours - it is probably more to do with quality of light and the nature of the environment and colours of the countryside."

For the past five years, Nobbs has been working closely with Tetsuya Sato of the Kyoto Institute of Technology, who has visited Leeds and Derby universities to help set up a project researching how the numerical expression of colour can take account of colour emotion.

The experiments take the form of semantic differencing techniques whereby volunteers selected for colour vision are shown a panel of 112 colours and several opposite pairs of adjectives such as vivid/dull, masculine/ feminine, striking/weak, cool/warm. They tick the box that they think best suits the colour. One of the most glaring differences concerned the use of passive or dynamic: the British sample group described some shades of greenish blue and bluish reds and purples as dynamic, while the Japanese decided these were passive. Similarly a number of blues, reds and purples were judged cool by 85 per cent of Japanese observers whereas 70 per cent of UK volunteers called them warm.

Nobbs has been surprised to discover so many discrepancies. "We have a lightness scale set up by Northern European observers in the 1930s, and they had the cheek to say that was how the rest of the world saw colour when it was not. When we look at a shade of blue or yellow and say they look equally as light, we are making an aesthetic judgement - if you set up the same scale with Japanese observers they would say they are not equally light. We've imposed our view on other cultures.

"The aesthetics of colour are treated much more seriously in Japan than here. There, scientists have been setting up these sorts of scales for years, but they found they could not get their publications accepted in scientific journals outside Japan."

It was largely this frustration that led to the collaboration between Leeds and Kyoto; Nobbs saw that the international scientific community was being slow to accept that colour can no longer be measured simply and objectively in terms of physics.

"So I volunteered to help. It was something I was once rather sceptical about myself," Nobbs admits. At first, his lectures on the subject were greeted with disparaging remarks such as "meaningless", "complete rubbish" or "not scientifically repeatable". But that is no longer the case, he says. There are now many who recognise the many benefits - economic and medical - to be derived from research in this area. The packaging industry, which relies on subliminal messaging, is an obvious beneficiary. Manufacturers of computer games are keen to know which colours best convey particular atmospheres, such as brutal, exciting, dramatic or calm, and the pharmaceuticals industry wants to understand whether people associate the colour of a pill or medicine with health-giving properties.

Sato believes the research will contribute to a greater understanding of cognitive science. He says that one of the most surprising results of his research was noticing how colour is used as an information and communication tool, for example in horse-race betting. "If jockeys do not wear colourful suits, it is difficult to know which horse won. Gamblers need to know that to get their money, so colour is used as a tool for communication through human cognition. If the gambler won, he or she would be happy, thus colour induces human feeling in the same way a writer communicates feelings to readers through words.

"I do not believe that when you are researching human feelings, science is the most useful tool. Science is useful for objective but not subjective analysis. It is better to approach research into sensations simply by using our eyes."

Sato points out that numerical expressions, which are useful for cross-cultural comparisons, are just one approach: "In my joint project with Jim and others, I am trying to derive empirical colour emotion formulae. It would be possible to develop an instrumental measuring system for the aesthetic properties of colours because mathematicians and physicists can easily make equations of algorithms corresponding to the aesthetic properties.

"But I think we need several concepts to analyse human feelings. I estimate that measuring the aesthetic properties of colours will have to wait until the numerical expression of human consciousness and the analysis of the higher stage of the human brain has taken place."

Nobbs thinks it may never be possible to have one world scale of colour measurement. "The best may be to go from one scale to another through a common system."